Configuration issue with Spark example

The example fails in two places when I try to run it on the 5.2 VM. I'm able to clone kite-examples, run mvn install, and run CreateEvents successfully.

```
mvn exec:java -Dexec.mainClass="org.kitesdk.examples.spark.CreateEvents"
```

I attempt to run the Spark job using this command.

```
spark-submit -class org.kitesdk.examples.spark.CorrelateEvents --jars $(mvn dependency:build-classpath | grep -v '^[' | sed -e 's/:/,/g') target/kite-spark-demo*.jar
```

1. The first time through, the Maven build is unable to find CorrelateEvents.class. Even though it fails, the class is built and it does appear in the target JAR.

```
Warning: Local jar /home/cloudera/kite-examples/spark/Downloading does not exist, skipping.
java.lang.ClassNotFoundException: org.kitesdk.examples.spark.CorrelateEvents
```

Running the command a second time, it fails with a permissions error.

```
Exception in thread "main" org.apache.hadoop.security.AccessControlException: Permission denied: user=cloudera, access=EXECUTE, inode="/user/spark":spark:spark:drwxr-x---
```

I searched for `inode="/user/spark"` and found this post on the Cloudera Community:

http://community.cloudera.com/t5/Advanced-Analytics-Apache-Spark/5-1-Quickstart-VM-issue/td-p/16684

It says that the permissions issue can be overcome by running the following commands.

```
$ sudo su - hdfs
$ hadoop fs -chmod -R 777 /user/spark
```

After running those commands, I was able to successfully run the example. I don't know how this issue inadvertently resolved itself when I ran the example for witnesses on February 17. This is probably why the issue wasn't caught before. Somewhere, running something, the permissions are set. When setting up the fresh VM, I started both Hue and Cloudera Manager with all services running.

So the question is, do we want to tell people to do the -chmod, then run the example twice, or should we adjust the example to handle both of these issues automagically?